J Mol Neurosci (1991) 3:!9-27
Journal of Molecular Neuroscience © Birkh~iuser Boston 1991
Role of Calcium in Regulation of Phosphoinositide Signaling Pathway Jitendra Patel, Richard A. Keith, Andre I. Salama, and W. Craig M o o r e ICI Pharmaceuticals Group, ICl Americas Inc., Concord Pike and Murphy Road, Wilmington, DE 19897, USA
Abstract. Using primary neuronal cultures we have examined the role of extracellular Ca 2+ in a receptorregulated phosphoinositide turnover. We report that rec e p t o r (glutamic a c i d and a c e t y l c h o l i n e ) - a c t i v a t e d phosphoinositide turnover requires the presence of extracellular Ca z+ (ECso = 21.1 p.M). The requirement for C a 2+ appears to be at an intracellular level and is highly selective for Ca 2 +. We also found that several inorganic and organic Ca z + channel blockers, including La 3 + and verapamil, inhibit phosphoinositide turnover. However, the pharmacological profile of these agents in this regard was distinct from their actions at the voltage-sensitive Ca 2 + channels. To explain the above requirement for extracellular Ca 2+ in agonist-stimulated phosphoinositide turnover and its sensitivity to Ca 2 +-channel blockers, we propose a hypothetical model suggesting that Ca 2+, following IP-3-mediated mobilization, exerts a facilitatory
action on the activity of receptor-phospholipase C complex. We further propose that in the absence of extracellular Ca2+ or in the presence of certain Ca 2+-channel blockers, refilling of calciosomes is ineffectual or inhibited, causing its depletion and subsequent inactivation of agonist-stimulated phosphoinositide turnover.
Many hormones, neurotransmitters and trophic factors mediate their actions via phospholipase Ccatalyzed hydrolysis of phosphatidylinositol 4,5bisphosphate. This reaction produces two intracellular messengers, inositol trisphosphate (IP-3) and diacylglycerol (DAG). IP-3 stimulates the mobilization of calcium from intracellular calcium ([Ca 2+ ]i) stores and DAG activates protein kinase C (Berridge and Irvine, 1989). Such receptor-mediated activation of phospholipase C is widely believed to be mediated via G-protein (Berridge and Irvine, I989). In addition to receptor-mediated activation, phospholipase C activity in excitable tissues also can be stimulated by activation of sodium channels (Gus-
Offprint requests to: J. Patel
ovsky et al., 1987) or membrane depolarization (Kendall and Nahorski, 1985). Such activation is thought to be mediated by an increase in [Ca 2+ ]i as a result of activation of voltage-sensitive calcium channels (VSCC) and is therefore sensitive to inhibition by VSCC antagonist (Zernig et al., 1986; Kendall and Nahorski, 1985; Gurwitz and SokoIovsky, 1987). In excitable tissue, therefore, the activation of phospholipase C can be either receptor-mediated or C a 2 + - m e d i a t e d (Eberhard and Holz, 1988; Crews et al., 1988). Receptor-activated PI turnover is generally considered to be independent of extracellular calcium ( [ C a 2 + ] o ) and its activity is accordingly largely unaffected by removal of [ C a 2 + ] o (Berridge, 1987). However, a number of recent studies have reported that agonist-stimulated PI turnover is sensitive to [Ca2+]o (see Discussion). Although these observations do not challenge the firmly established belief that the rise in [Ca2+]i is a consequence, and not a forerunner, of PI turnover, the role of [Ca 2 + ]o in the regulation of PI turnover has not been explained satisfactorily. Here we report that agoniststimulated PI turnover is prevented if extracellular Ca 2+ is either absent or influx of C a 2+ is inhibited by a Ca 2+ channel blocker. To explain the Ca z+ dependency, we suggest that Ca 2+ released from calciosomes acts, via an unknown mechanism, to sustain phospholipase C activity.
Materials and Methods [3H]Inositol was purchased from Amersham. Quisqualic acid, carbachol, and verapamil were from Research Biochemicals. DNQX was supplied by Tocris Neuramin. AG l-X8 was from Bio-Rad Laboratories. PN 200-110 was a kind gift from Sandoz. ~o-Conotoxin was obtained from Bachem. All other drugs were obtained from standard scientific chemical suppliers.
20
Patel et al.: Role of Ca 2+ in Regulation of PI Turnover
Cell culture Primary cultures of rat cortical neurons were prepared based on a procedure described previously (Dichter, 1978). Briefly, cells from cortices of fetal rats (Sprague-Dawley, 17-19 days' gestation) were dispersed by brief trypsinization and gentle trituration. Cells then were plated at a density of 1.4 × 105 cells/cm 2 in culture plates precoated with polyL-lysine (Sigma) in plating medium containing glutamine-free Eagle's minimal essential medium (Gibco), 10% fetal bovine serum, and 10% heatinactivated equine serum (HyClone). Proliferation of nonneuronal cells was inhibited at confluency by two-day treatment with 10 ixM cytosine-D-arabinofuranoside (Sigma). After seven days in vitro, the culture medium was switched to 10% Eagle's minimum essential medium with 10% heat-inactivated equine serum. Receptor-activated [3H]inositol phosphate production Receptor-activated PI turnover was measured essentially as described previously (Sladeczek et al., 1988). Briefly, cultures of 13-15 days were prelab e l e d by o v e r n i g h t i n c u b a t i o n of 2 p~Ci/ml [3H]inositol. Before the experiment, the culture medium was replaced with a controlled salt solution (CSS) containing 20 mM HEPES, 115 mM NaCI, 12 mM KCi, 2.5 mM CaCl2 and 10 mM LiCl (pH 7.4). Drugs, when evaluated, were preincubated with the cells for I0 minutes. Calcium-free condition was obtained by omitting calcium and including 30 ~M EGTA in CSS. PI turnover was initiated by the addition of specified agonist at indicated concentration and was terminated after 30 minutes by the addition of 4.5% (final) perchloric acid. All the above procedures were performed at 37°C. Following neutralization of the cell extracts, inositol phosphates were isolated collectively or individually by ion-exchange chromatography as described previously (Downes et al., 1982).
Results As previously reported, muscarinic and glutaminergic receptor agonists, carbachol (CB), and quisqualate (QA), respectively, are potent activators of PI turnover. However, in the absence of [Ca2÷]o both QA (Fig. 1) and CB (Fig. 2) gave no stimulation of PI turnover as measured either by total inositol phosphates or IP-3 production (Fig. 3). This effect of calcium was reversible as reintroduction of Ca 2 + into the incubation medium after preincubation with EGTA restored the response to QA (Fig.
I00,
,
~ z_
-zo+
+
+
+
+
+
DHOX
+
I OOflM Q A +
2_SmM
TTX
Co ++
B a ++ O T H E R
ADDITIONS
Fig. I. Modulation of QA-stimulated PI turnover: requirement for extracellular calcium. Production of inositol phosphates in response to QA (100 nM) was measured in primary neurocortical cultures in the absence and presence of extracellular Ca 2+, and with cells preincubated (10 minutes) in Ca2+-free media. Effects of DNQX, TTX and replacement of extracellular Ca2+ with Ba2÷ on QAstimulated inositol phosphates production are also shown. The inositol phosphate production is expressed as percentage of QA response obtained in the presence of extracellular Ca~+ 1). Thus, removal of [Ca2÷]o is unlikely to have severely compromised cellular integrity or to have caused dephosphorylation of PIP-2 to PI (see Brammer and Weaver, 1989). The requirement for Ca 2 ÷ appears to be highly specific. Thus, in accordance with a previous report (Eberhard and Holz, 1987), Ca 2÷ could not be substituted by Ba 2÷ (Figs. 1 and 2). This level of specificity would argue against a Ca 2÷ requirement solely for binding to polyphosphinositides, as sug-
--I'I00
cz u~ L~
40
~
20
|
o
I
-zu,
.
.
+
+
.I.
+ TTX
I -
.i.
. ,i-
Be
ImM
CB
2.SmM OTHER
Ca ADOITIONS
Fig. 2. Modulation of CB-stimulated PI turnover: requirement for extracellular Ca 2+. CB-stimulated inositol phosphate production was measured in neurocortical cultures in the absence of extracellular Ca 2+, where extracellular Ca2+ was replaced by Ba2+, and in the presence TTX. The production of inositol phosphates is expressed as percentage of that obtained with 1 mM CB in the presence of 2.5 mM CaC1.
Patel et al.:
Role of Ca 2+ in Regulation of PI Turnover
400
A:
I"IEHBRAN£IMPERI'IE&IIT BAPT&
21 B: I'IIEI'IOR&HEPERMEANTIBAPTA
~30'
i
0
7-
300
mmZO,
I
10,
0 n-
100
0 z
+
-
-
-
+
+
+
+
-
+ -
-
lOOnM +
-
I mM
2.SmM
QA CO
Ca
Fig. 3. Effect of removing extracellular C a 2 + on QA- and CB-stimulated IP-3 synthesis. 100 nM QA- and l mM CB-stimulated IP-3 synthesis in neurocortical cultures was measured in the absence and presence of 2.5 mM C a 2 + . The IP-3 synthesis was measured as described in the text. gested before (Hendrickson and Reinertsen, 1969). The dose-response relationship for Ca 2 + is shown in Fig. 4 and demonstrates an ECs0 of 21.1 p.M. A similar calcium requirement has been reported previously for histamine-stimulated PI turnover (Kendall and Nahorski, 1984). To establish whether the Ca 2+ requirement was intra- or extracellular, cells preloaded with BAPTA to preferentially chelate [Ca2+]i were used. Under such conditions, both QA- and CB-stimulated PI turnover was abolished (Fig. 5). Influx of calcium through VSCCs has been 1101009080,~ 70 tit/"
~ 0o
//
~ 5o 0
/
40-
/
/
/
/
/
30.~///
2010I
1.0
t
I
i
10 100 1000 Extracellular Calcium (I~M)
I
10000
Fig. 4. Dose-response relationship for C a 2 + requirement for QA-PI turnover. The 100 nM QA-stimulated total inositol phosphate production in neurocortical cultures was measured as described in the text at various concentrations of extracellular Ca 2+. The curve is a computergenerated nonlinear least squares fit of data to the logistic equation.
Fig. 5. Effect of chelating intracellular C a 2 + on QA- and CB-stimulated PI turnover. Primary neuronal cultures were preincubated for 10 minutes with 50 w.M membrane impermeant (A), and membrane permeant (B) form of BAPTA. The treated cells (shaded bars) and untreated cells then were used for measuring 100 nM QA- and I mM CB-stimulated inositol phosphates production over a 30minute period. widely observed to stimulate PI turnover (Eberhard and Holz, 1988). QA via a specific ionotropic receptor can cause depolarization of membranes and thereby activate calcium influx via VSCC (Murphy and Miller, 1989). However, it has been established previously (Patel et al., 1990; Sladeczek et al., 1988) that QA-stimulated PI turnover is not elicited via this pathway and accordingly is insensitive to DNQX, an antagonist of the ionotropic action of QA. Nevertheless we examined the effect of VSCC blockers on PI turnover. VSCC can be effectively blocked by La 3+, Cd 2+ , Ni 2+ , and Co 2+ (with respective ICs0 = 0.9, 7.0, 280, and 560 ~M) (Narahashi et al., 1987). These cations inhibited QA-stimulated PI turnover with the following ICso (~M): La 3+, 6.55; Cd 2+, 141; Ni 2+ , 655; Co 2+ , 1530 (Fig. 6). Three different classes of organic VSCC blockers, viz. verapamil, diltiazem, and PN 200-110, were tested also. Verapamil inhibited QA-stimulated PI turnover with an ICso of 178 ~M (Fig. 7). The inhibitory effect of verapamil was noncompetitive with respect to agonist (suggesting that its effect was not due to displacement of agonist from its receptor). In contrast to the highly stereoselective action of verapamil on the VSCC, its inhibition of PI turnover was not stereoselective (data not shown). Diltiazem (Fig. 7) and PN 200-110 also inhibited QA-stimulated PI turnover; however, PN 200-110 only produced a partial inhibition (up to 50%, data not shown). Relatively high concentrations of VSCC antagonists were required to block PI turnover. Although this may suggest a nonspecific mechanism of action, the structural diversity of these agents argues against such a conclusion.
22
Patel et al.: Role of Ca2÷ in Regulation of PI Turnover
0o:00
100
,o t
..................I~..
90 :
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
~
80
,,\
'\
c
70
m 60 50
"\ ~
..o
',,
'.R g
\
O '~Co 2+
\!,,
\,
0
Journal of Molecular Neuroscience © Birkh~iuser Boston 1991
Role of Calcium in Regulation of Phosphoinositide Signaling Pathway Jitendra Patel, Richard A. Keith, Andre I. Salama, and W. Craig M o o r e ICI Pharmaceuticals Group, ICl Americas Inc., Concord Pike and Murphy Road, Wilmington, DE 19897, USA
Abstract. Using primary neuronal cultures we have examined the role of extracellular Ca 2+ in a receptorregulated phosphoinositide turnover. We report that rec e p t o r (glutamic a c i d and a c e t y l c h o l i n e ) - a c t i v a t e d phosphoinositide turnover requires the presence of extracellular Ca z+ (ECso = 21.1 p.M). The requirement for C a 2+ appears to be at an intracellular level and is highly selective for Ca 2 +. We also found that several inorganic and organic Ca z + channel blockers, including La 3 + and verapamil, inhibit phosphoinositide turnover. However, the pharmacological profile of these agents in this regard was distinct from their actions at the voltage-sensitive Ca 2 + channels. To explain the above requirement for extracellular Ca 2+ in agonist-stimulated phosphoinositide turnover and its sensitivity to Ca 2 +-channel blockers, we propose a hypothetical model suggesting that Ca 2+, following IP-3-mediated mobilization, exerts a facilitatory
action on the activity of receptor-phospholipase C complex. We further propose that in the absence of extracellular Ca2+ or in the presence of certain Ca 2+-channel blockers, refilling of calciosomes is ineffectual or inhibited, causing its depletion and subsequent inactivation of agonist-stimulated phosphoinositide turnover.
Many hormones, neurotransmitters and trophic factors mediate their actions via phospholipase Ccatalyzed hydrolysis of phosphatidylinositol 4,5bisphosphate. This reaction produces two intracellular messengers, inositol trisphosphate (IP-3) and diacylglycerol (DAG). IP-3 stimulates the mobilization of calcium from intracellular calcium ([Ca 2+ ]i) stores and DAG activates protein kinase C (Berridge and Irvine, 1989). Such receptor-mediated activation of phospholipase C is widely believed to be mediated via G-protein (Berridge and Irvine, I989). In addition to receptor-mediated activation, phospholipase C activity in excitable tissues also can be stimulated by activation of sodium channels (Gus-
Offprint requests to: J. Patel
ovsky et al., 1987) or membrane depolarization (Kendall and Nahorski, 1985). Such activation is thought to be mediated by an increase in [Ca 2+ ]i as a result of activation of voltage-sensitive calcium channels (VSCC) and is therefore sensitive to inhibition by VSCC antagonist (Zernig et al., 1986; Kendall and Nahorski, 1985; Gurwitz and SokoIovsky, 1987). In excitable tissue, therefore, the activation of phospholipase C can be either receptor-mediated or C a 2 + - m e d i a t e d (Eberhard and Holz, 1988; Crews et al., 1988). Receptor-activated PI turnover is generally considered to be independent of extracellular calcium ( [ C a 2 + ] o ) and its activity is accordingly largely unaffected by removal of [ C a 2 + ] o (Berridge, 1987). However, a number of recent studies have reported that agonist-stimulated PI turnover is sensitive to [Ca2+]o (see Discussion). Although these observations do not challenge the firmly established belief that the rise in [Ca2+]i is a consequence, and not a forerunner, of PI turnover, the role of [Ca 2 + ]o in the regulation of PI turnover has not been explained satisfactorily. Here we report that agoniststimulated PI turnover is prevented if extracellular Ca 2+ is either absent or influx of C a 2+ is inhibited by a Ca 2+ channel blocker. To explain the Ca z+ dependency, we suggest that Ca 2+ released from calciosomes acts, via an unknown mechanism, to sustain phospholipase C activity.
Materials and Methods [3H]Inositol was purchased from Amersham. Quisqualic acid, carbachol, and verapamil were from Research Biochemicals. DNQX was supplied by Tocris Neuramin. AG l-X8 was from Bio-Rad Laboratories. PN 200-110 was a kind gift from Sandoz. ~o-Conotoxin was obtained from Bachem. All other drugs were obtained from standard scientific chemical suppliers.
20
Patel et al.: Role of Ca 2+ in Regulation of PI Turnover
Cell culture Primary cultures of rat cortical neurons were prepared based on a procedure described previously (Dichter, 1978). Briefly, cells from cortices of fetal rats (Sprague-Dawley, 17-19 days' gestation) were dispersed by brief trypsinization and gentle trituration. Cells then were plated at a density of 1.4 × 105 cells/cm 2 in culture plates precoated with polyL-lysine (Sigma) in plating medium containing glutamine-free Eagle's minimal essential medium (Gibco), 10% fetal bovine serum, and 10% heatinactivated equine serum (HyClone). Proliferation of nonneuronal cells was inhibited at confluency by two-day treatment with 10 ixM cytosine-D-arabinofuranoside (Sigma). After seven days in vitro, the culture medium was switched to 10% Eagle's minimum essential medium with 10% heat-inactivated equine serum. Receptor-activated [3H]inositol phosphate production Receptor-activated PI turnover was measured essentially as described previously (Sladeczek et al., 1988). Briefly, cultures of 13-15 days were prelab e l e d by o v e r n i g h t i n c u b a t i o n of 2 p~Ci/ml [3H]inositol. Before the experiment, the culture medium was replaced with a controlled salt solution (CSS) containing 20 mM HEPES, 115 mM NaCI, 12 mM KCi, 2.5 mM CaCl2 and 10 mM LiCl (pH 7.4). Drugs, when evaluated, were preincubated with the cells for I0 minutes. Calcium-free condition was obtained by omitting calcium and including 30 ~M EGTA in CSS. PI turnover was initiated by the addition of specified agonist at indicated concentration and was terminated after 30 minutes by the addition of 4.5% (final) perchloric acid. All the above procedures were performed at 37°C. Following neutralization of the cell extracts, inositol phosphates were isolated collectively or individually by ion-exchange chromatography as described previously (Downes et al., 1982).
Results As previously reported, muscarinic and glutaminergic receptor agonists, carbachol (CB), and quisqualate (QA), respectively, are potent activators of PI turnover. However, in the absence of [Ca2÷]o both QA (Fig. 1) and CB (Fig. 2) gave no stimulation of PI turnover as measured either by total inositol phosphates or IP-3 production (Fig. 3). This effect of calcium was reversible as reintroduction of Ca 2 + into the incubation medium after preincubation with EGTA restored the response to QA (Fig.
I00,
,
~ z_
-zo+
+
+
+
+
+
DHOX
+
I OOflM Q A +
2_SmM
TTX
Co ++
B a ++ O T H E R
ADDITIONS
Fig. I. Modulation of QA-stimulated PI turnover: requirement for extracellular calcium. Production of inositol phosphates in response to QA (100 nM) was measured in primary neurocortical cultures in the absence and presence of extracellular Ca 2+, and with cells preincubated (10 minutes) in Ca2+-free media. Effects of DNQX, TTX and replacement of extracellular Ca2+ with Ba2÷ on QAstimulated inositol phosphates production are also shown. The inositol phosphate production is expressed as percentage of QA response obtained in the presence of extracellular Ca~+ 1). Thus, removal of [Ca2÷]o is unlikely to have severely compromised cellular integrity or to have caused dephosphorylation of PIP-2 to PI (see Brammer and Weaver, 1989). The requirement for Ca 2 ÷ appears to be highly specific. Thus, in accordance with a previous report (Eberhard and Holz, 1987), Ca 2÷ could not be substituted by Ba 2÷ (Figs. 1 and 2). This level of specificity would argue against a Ca 2÷ requirement solely for binding to polyphosphinositides, as sug-
--I'I00
cz u~ L~
40
~
20
|
o
I
-zu,
.
.
+
+
.I.
+ TTX
I -
.i.
. ,i-
Be
ImM
CB
2.SmM OTHER
Ca ADOITIONS
Fig. 2. Modulation of CB-stimulated PI turnover: requirement for extracellular Ca 2+. CB-stimulated inositol phosphate production was measured in neurocortical cultures in the absence of extracellular Ca 2+, where extracellular Ca2+ was replaced by Ba2+, and in the presence TTX. The production of inositol phosphates is expressed as percentage of that obtained with 1 mM CB in the presence of 2.5 mM CaC1.
Patel et al.:
Role of Ca 2+ in Regulation of PI Turnover
400
A:
I"IEHBRAN£IMPERI'IE&IIT BAPT&
21 B: I'IIEI'IOR&HEPERMEANTIBAPTA
~30'
i
0
7-
300
mmZO,
I
10,
0 n-
100
0 z
+
-
-
-
+
+
+
+
-
+ -
-
lOOnM +
-
I mM
2.SmM
QA CO
Ca
Fig. 3. Effect of removing extracellular C a 2 + on QA- and CB-stimulated IP-3 synthesis. 100 nM QA- and l mM CB-stimulated IP-3 synthesis in neurocortical cultures was measured in the absence and presence of 2.5 mM C a 2 + . The IP-3 synthesis was measured as described in the text. gested before (Hendrickson and Reinertsen, 1969). The dose-response relationship for Ca 2 + is shown in Fig. 4 and demonstrates an ECs0 of 21.1 p.M. A similar calcium requirement has been reported previously for histamine-stimulated PI turnover (Kendall and Nahorski, 1984). To establish whether the Ca 2+ requirement was intra- or extracellular, cells preloaded with BAPTA to preferentially chelate [Ca2+]i were used. Under such conditions, both QA- and CB-stimulated PI turnover was abolished (Fig. 5). Influx of calcium through VSCCs has been 1101009080,~ 70 tit/"
~ 0o
//
~ 5o 0
/
40-
/
/
/
/
/
30.~///
2010I
1.0
t
I
i
10 100 1000 Extracellular Calcium (I~M)
I
10000
Fig. 4. Dose-response relationship for C a 2 + requirement for QA-PI turnover. The 100 nM QA-stimulated total inositol phosphate production in neurocortical cultures was measured as described in the text at various concentrations of extracellular Ca 2+. The curve is a computergenerated nonlinear least squares fit of data to the logistic equation.
Fig. 5. Effect of chelating intracellular C a 2 + on QA- and CB-stimulated PI turnover. Primary neuronal cultures were preincubated for 10 minutes with 50 w.M membrane impermeant (A), and membrane permeant (B) form of BAPTA. The treated cells (shaded bars) and untreated cells then were used for measuring 100 nM QA- and I mM CB-stimulated inositol phosphates production over a 30minute period. widely observed to stimulate PI turnover (Eberhard and Holz, 1988). QA via a specific ionotropic receptor can cause depolarization of membranes and thereby activate calcium influx via VSCC (Murphy and Miller, 1989). However, it has been established previously (Patel et al., 1990; Sladeczek et al., 1988) that QA-stimulated PI turnover is not elicited via this pathway and accordingly is insensitive to DNQX, an antagonist of the ionotropic action of QA. Nevertheless we examined the effect of VSCC blockers on PI turnover. VSCC can be effectively blocked by La 3+, Cd 2+ , Ni 2+ , and Co 2+ (with respective ICs0 = 0.9, 7.0, 280, and 560 ~M) (Narahashi et al., 1987). These cations inhibited QA-stimulated PI turnover with the following ICso (~M): La 3+, 6.55; Cd 2+, 141; Ni 2+ , 655; Co 2+ , 1530 (Fig. 6). Three different classes of organic VSCC blockers, viz. verapamil, diltiazem, and PN 200-110, were tested also. Verapamil inhibited QA-stimulated PI turnover with an ICso of 178 ~M (Fig. 7). The inhibitory effect of verapamil was noncompetitive with respect to agonist (suggesting that its effect was not due to displacement of agonist from its receptor). In contrast to the highly stereoselective action of verapamil on the VSCC, its inhibition of PI turnover was not stereoselective (data not shown). Diltiazem (Fig. 7) and PN 200-110 also inhibited QA-stimulated PI turnover; however, PN 200-110 only produced a partial inhibition (up to 50%, data not shown). Relatively high concentrations of VSCC antagonists were required to block PI turnover. Although this may suggest a nonspecific mechanism of action, the structural diversity of these agents argues against such a conclusion.
22
Patel et al.: Role of Ca2÷ in Regulation of PI Turnover
0o:00
100
,o t
..................I~..
90 :
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
~
80
,,\
'\
c
70
m 60 50
"\ ~
..o
',,
'.R g
\
O '~Co 2+
\!,,
\,
0
